Summary
As part of the reform of railroad legislation in Germany, the introduction of efficiency-oriented incentive regulation has been proposed for railroad access charges. One major objective of incentive regulation is to provide railway infrastructure providers sufficient incentives for cost savings and to make sure that access charges reflect productivity improvements appropriately. One key aspect of incentive regulation is the determination of the sectoral productivity growth. Against this background, the study examines how productivity growth in the railroad sector can be determined in Germany.
The study describes and compares methods for determining productivity factors in the regulatory practice in other infrastructure sectors and/or countries. The examples show that a variety of methods is used to determine productivity factors. More sophisticated benchmarking-methods (e.g. Stochastic Frontier Analysis, Data Envelopment Analysis) are mainly applied for individual, firm-specific inefficiencies and require extensive data sets (and cannot be applied if such data are not available). In addition to assessing inefficiencies directly, cost models and cost audits are often used to identify cost drivers and thus enable and inform regulators to collect data more meaningfully for benchmarks in subsequent regulatory periods. General productivity growth in the sector can be estimated based on benchmarks with a ‘synthetic comparator sector’. Advantages of this method are high transparency and the easy applicability. A shortcoming of this method is that reasons for changes in productivity cannot be identified. By using mainly competitive sectors as comparator sectors, the productivity index primarily approximates technological progress (and does not capture catch-up effects due to the elimination of inefficiencies).
The study goes on to calculate productivity growth in the German railroad sector based on a synthetic comparator sector. Our calculations estimate an average annual increase in productivity of 0.6 % between 1992 and 2008. The results of different scenarios vary between 0.26 % and 1.85 % p.a. These values should be seen as lower bounds for the determination of regulatory productivity targets. For the first periods of an incentive regulation, a catch-up factor should be taken into account separately to reflect the reduction of inefficiencies that had likely been accumulated before incentive regulation.
Generally, we recommend that productivity targets should be determined using a pragmatic approach. The simpler, the more transparent and more understandable the methodologies and analyses used to support regulatory decisions are, the more accepted will be the results.
Discussion Paper is available for download.
